Stream Processing for Real-time Intelligent Sensor Data Processing
Reference persons FULVIO CORNO
Research Groups E-LEARNING, E-INTELLIGENCE, E-INTERACTION GROUP - E-LITE
Thesis type APPLIED RESEARCH
Description Exploit the novel "Stream processing" database technologies for computing, in real-time, user-friendly results form streams of sensor data.
Several applications in Smart Buildings require the ability to perform complex processing, in real-time, starting from hundreds of sensors disseminated in the environment. Such sensors may monitor power consumption, temperature, humidity, human presence, device statuses (on/off, open/closed), external weather, air/water flows, ... Energy Managers, and other Building Managers, need to rely on a simple, intuitive, real-time display showing aggregate and post-processed information.
EsperTech Stream processing (or Complex Event Processing) is a novel paradigm for data management and processing, where a pre-defined set of queries is applied to a memory-less, continuous-flowing stream of events from different sources. One reliable and efficient open source library for complex event processing is Esper (from EsperTech).
The purpose of the thesis is to provide a high-level framework, based on Stream Processing Blocks, that enables energy managers to define their own computational "chains". Such blocks must be easily understandable by the domain expert (energy engineer), with no programming background. The sofware should then automatically translate the specification into the low-level Esper queries.
A preliminary version of this translation software is already available, and the thesis will concentrate more on high-level block definition, and on user interfaces for creating processing chains and for visualizing results.
Required skills Java, Data bases
Deadline 11/11/2012 PROPONI LA TUA CANDIDATURA